When unmanned systems operate autonomously, it is important to predict and evaluate the operational safety of the evaluation system. The evaluation of operational safety depends on the method of predicting the operating state of unmanned systems. However, short time prediction methods based on physical drive or partial data drive can only ensure the security of the operating state of the unmanned system during operation, but cannot guarantee that the operation will not exceed the predetermined safety area from a global perspective. In order to solve this problem, this paper proposes a long time prediction method for operational safety based on path planning. A* algorithm is used for path planning, and the operation safety probability expression of unmanned system is obtained through repeated planning, and then the operation safety of unmanned system is quantitatively evaluated. The effectiveness of the proposed method is verified by experiments. In addition, this global prediction method can be deployed by the airborne computing platform to realize the planning and confirmation of the safe operation path of the unmanned system, so as to ensure the autonomous and safe operation of the unmanned system.

Unmanned systems,Operational safety prediction,A* algorithm